Use of complete Fruend's adjuvant for production of antibodies to study protein conformation is valid only if emulsification in adjuvant does not denature protein antigens. Using electron paramagnetic resonance to observe directly the protein in situ in the opaque emulsions, we demonstrate that hemoglobin is not denatured by emulsification or storage in adjuvant for 24 hours at 4 degrees C. The results also rule out one proposed mechanism of adjuvant action, via alteration of protein structure, at least in this case.